 Well, thank you everyone for, okay, no, it's good. Hi, that's better. Well, good afternoon, everybody. My name's Carly Green. I'm part of the panel, but today, right now, I'm going to be representing GFY to open this session. Thanks for coming along to learn more about the IPCC software. And we have a panel today led by Sandra Federici from the IPCC technical support unit. Myself and Inga will present on the wall to wall add-on that we've been working on. And then we also have Alfonso and Giulio to talk about the sample-based add-on as well, the collect Earth add-on. And yeah, we really hope you enjoy the session and look forward to your interaction over the next hour or so. I'm going to hand over to you now, Sandra. Thanks. Okay, thank you very much. Thanks for coming. Let me see how it works. Where should I point? Okay, this way is better. So my presentation in terms of slide is quite long, so I'm not going to present you each and every slide, but the goal was to eat a kind of self-standing presentation that people may revisit at home. What I'm going to present is, first, the IPCC inventory software in general. And then an example can be used for instance, for Red Plus. I choose Red Plus because of the, I mean, you are in GFI. That has been very active on Red Plus, but of course can be used for any land-related estimation. So let's go with the software. This is the software. The mask is very basic. All the IPCC categories on the left-hand side. By select, you have all the sectors. This is not just for land. It's not just for forest. It's not just for a follow. It's all sectors. You select on the left the category for which you want to prepare the estimates. You click it, and then on the main window, you have all the worksheets for the calculation of the estimate associated with that. Category. This is, again, this is energy. Energy is easy. The ZF is much more complicated. Energy with two worksheets. You have the three IPCC tier methodological approaches that you may implement in a combination of different subdivisions. So the software is for national gas inventories, but can be applied to subnational level. You may compile an infinite number of projects, regions, cities, whatever you want. Up to the total. This is the architecture of the software. It has functionality in terms of administration. Yes. He's... What can I do? Sorry. Julio does it. It's okay. But now it's not okay there. Maybe it's because of the Zoom. Zoom, I need to... Wait a moment. Share screen. Now I guess it's okay. Now it's okay also for Zoom. What I have to do. Still, it's not in... Sorry. Now I have this one there. You have to double the screen instead of double it. What is it? It's the same level. Oh, sorry. Anyway, in the meanwhile, these are the building blocks of the software. There are administration functions. You select the country. You select the users. You may have a multiple user on the same database. You select the years of the... An inventory is a time series of inventory years. You select the time series of years. You... Okay. I guess I can go. Then you have the wall sheets for data entry, which is the main element where you input your data according with the data input in the wall sheets. You have the calculation of emissions and removals. You have a functionalities for queuing the sea and certain analysis, key category analysis. You have for the energy sector, the reference method. The software contains all, besides some potential errors, default data. It has data managers to make easier the job of the compiler. We are building data managers. And you see most of the data managers are for the... They are for the sector because it's the one more demanding in terms of data input. So there is one... There is a fuel manager. It's not for us, but there is a land representation manager. There is a land use type managers. And there is a livestock manager, a waste type manager. The software is a database. So all data are archived there. There is the possibility for the user to add an infinite number of comments and notes in his calculation wall sheets, in the estimates produced. So everything is stored. This makes a warrant for the continuation of inventory when you have different people that alternate in the software preparation. Yes, functionalities for data export and import. We are extending this functionality, we'll see, to make the software interoperable with the UNFCCC reporting tool for the National Greenhouse Gas Inventories. In practice, the software is a Microsoft Access database. These are some pluses and cons. One cons is that this software doesn't run on an Apple computer. It's not going to run for the next time. Sooner or later, maybe we will do something better, this is a limitation that you need to keep. Now it doesn't move anymore. There is a technical assistance. What should I do? It doesn't... I found a different function. So, as I said, the software now is considered one of the central elements in the ETF under the UNFCCC because it allows to prepare an inventory but also to export the information into the common reporting tables of UNFCCC. So, they function that this version has not, but the next version that we will release next month, the UNFCCC subsidiary bodies, will have the functionality to then populate the CRT tables with the information and the estimates prepared by the software. By June it will be for the energy sector, then there will be waste, and then hopefully by next couple, there will be a follow, and then last EPU sector. One of the things, one of the powerful elements of the software again is the possibility to archive all the information. That's why we are pushing also developed countries to use it as an instrument to ensure that archiving also has a comparison with their estimates. So, it can be a kind of default estimates that can be compared with their own. The software implement all IPCC tiers and approaches, all. In all sectors, there is a lot of flexibility there. There are a large number, and I will see, of worship calculation, but it doesn't mean that you need to compile each and every worship to have your estimates, but you may accommodate your approach into the software very likely. If you have a problem, we are happy to interact to see if something is to that, or just to provide the advice on how to use the software according to your dataset. Provide all the IPCCs default, and of course allow to input your user-specific data. Allow multiple users to work on the software. The software can be a database, our suggestions to put the database in the cloud with a super user. The super user allow, give the access to section of the database to different actors, according with their expertise. He's already implemented the AR5 with a warming potential of 100 years, which is the requirement. Paris Agreement, there is all the wetland supplement methodological elements are already in. They can be easily identified because of the color. It's a kind of lilac, then it depends on the computer, many things. We are still working to further develop the software. So in the next version, there will be also the possibility to calculate the indirect CO2 emissions, the indirect CO2 emissions. There will be all the interoperability done. There will be a revision of the notation key how they are currently used in order to make it easier and more powerful, let's say. I'm serious as now the current software has a big limitation in terms of the export import of the IPCC table and all of the CRTs. Just year by year, this is tedious. If you have a 20 years time series, you need to download the 20 files, which is, of course. We will fix this. We are also working on supporting tools. Two of them will be presented today. They are bought for land representation. They are under development by FEO. One is based on this PAL tool. Then Inge and Carly will present on that. The other one is based on the Colette-Heart tool. Danilo and Alfonso will present on that. This is how the interoperability is expected to work. You see, you have those blue are the IPCC inventory software washes. By starting the functionality, the software transfers the data according to the mapping. We will publish this huge work of mapping. Each and every cell of the CRTs is mapped into the software. The software transfers the data into the CRTs. You will see these CRTs visualized in the software. Then you may transfer everything by using a JSON file. Okay, I already said about the representation. There are also other two Excel-based tools that are relevant for a forum. One is the Arbex VOOP product that we developed together with FAUSE, with the Julia team, with the Francesco team. The other one is the land representation logic. The land representation is an Excel-based tool that is going to be developed by the office. But still, also this one is under development. They are complex things. Another very important element, unfortunately, because of this black bar here, I don't know if it's possible to move this, but probably no. Now, in this way, it's better, no? Even better if I put it on the side, no. Maybe for some slide, there will be a problem. Now you see, because this is very important, we are building guidebooks, sectoral guidebook, because the software is not rocket science, as I said, but it's not even so banal. I mean, it's needed to have a kind of instruction. The instruction is a guidebook that will guide the user step-by-step in the use of the IPCC worksheets and functionalities to prepare the estimates for each and every category. Now, I don't know. Anyhow, I don't know if I have any other representation. For the energy sector, the guidebook is already prepared. Of course, the energy sector is the most important. I follow his second in terms of the greenhouse gas emissions and removals, which is our activity. We hope to release the guidebook by the end of the year, because even if it's ready, we need somebody that takes some of this interest by his energy. We need some external reviewer that have a look and provide feedback before proceeding. Now, the second part of the presentation is about an example. I applied the IPCC inventory software to prepare estimates from red activities. I need to give you this big disclaimer. Read it very, very big, because IPCC has no position, so it didn't produce anything on red. So this is a very simple aspect, even banal interpretation on red, but I just done just to show the calculation, not to tell you how to do red. Now, the example is a territory with the three land use classes, forest land, the grass, and the cropland with a dynamic that is conversion of primary forest to secondary forest. Conversion of secondary forest to forest plantation and vice versa. Conversion of primary forest to cropland and conversion to grassland to forest plantation. These are the dynamic. And all these dynamic, which is here, these are the processes. But just for a minute, don't take this. But of course, the conversion of a manager forest land to annual generic cropland is at the forest station. The manager change of a manager natural primary forest to manager natural secondary forest is likely a long-term average carbon stock degradation. And then you have a degradation also in the conversion of natural forest to plantation. You have a carbon stock enhancement in the conversion of plantation to natural forest and a long-term enhancement of carbon stock in a conversion from manager to grassland to forest land. The sustainable manager forest is made with these characteristics that I simulate some changes. I prepared to that base. The software may manage an infinite number of databases, of course, even if it cannot visualize only one database at a time. So you switch from one database to another. You don't open to database at the same time. So you cannot do the comparison between two databases into the software. I did one database with a reference level, a reference level which is the historical period which is the gray. Then you have an actual period where you have the projection made in the reference level and then the actual estimates. And then you have another period that is just projected because in the future. We assume that you are in 2025. We are already in the future. The projection, of course, because I don't take any... I will be in neutral. This is the scheme. I just projected the historical rate. So the reference level is historical without any changes. You see, there are colors there that indicate, for instance, primary forests are subject to conservation, but if I consider primary forest, I reduce the deforestation, I reduce the forest degradation and so on. Many activities may be on the same unit or land, but it doesn't matter. So far as this is a national inventory, all the territory is reported, so everything is captured without the need to label each single piece. These are the data. These are the data of the actual land. These are the harvesting data, the project and the actual. The harvesting modelized with the age class, so no any changes, just business, pure business as usual. What do you need to do in the software? First, you need to compile the land use types in the land use manager. As shown here, we have a manager forest land, we have a manager forest land, we have forest plantation. You need to create this into the land use manager. I will show you one. You input the land representation, which is all the areas of the unit or land, it changes across time. Finally, you find the unit or land in the washes where you calculate the carbon stock, and then there you input any additional information that you need to calculate the carbon stock changes. The exercise is only about carbon. I didn't put anything else complicated. The land use manager type is by default in the 12 main land use subcategories. There are the six IPCC, and then we have a manager and manager for forest land, wetlands, other land and grassland. Then we have an annual perennial for cropland. This is because we have different methodological approaches according to this distinction, manager and manager, the annual perennial. For settlement, you have a tree, and let's say without perennial vegetation, which also in that case, give you different methodological approaches. This is the land use manager type. This is for a manager forest land, so there are not so many factors to input because you don't have emissions removal in a manager land use forest land, and the main manager forest land, because IPCC assume that these are mostly not anthropogenic. Be aware of what is important, but probably you don't see because the screen is not so big and the pointer doesn't work. But what is this? Each subdivision is soil type specific, climate type specific, vegetation zone specific, so you cannot switch from one, of course, from one land use subdivision that is on a high clay soil to another one that you find on a low clay soil. This is a big constraint. Also avoid the error, simplify also the compilation. Because when you have the, now I show you, this is the secondary forest, there are many more parameters. Black means the parameters are used for tier one, blue means that the parameters are used for tier two calculation. They are all the land use type. Soil, then you select the soil, not for each of these land use subdivision, but by default you have the IPCC, but if you want you build your own soil classification and the software works with your own soil classification. So in this case I added terra preta, a new soil type and do the calculation according with that. Same for the climate, you have the IPCC climate classification zones, but you may go for a different climate classification. In this case I put an Eastern Amazonian climate and then I may do the calculation according with this. Of course IPCC provide you the default according with the IPCC stratification. If you change soil classification, and or climate classification, you don't have any more the IPCC default offered by the software because those defaults may and may not apply. So you need to input your own estimates. What is a consistent land representation? Assume you know the IPCC software allow to input each data according to each and every land use approach of IPCC, approach one, no land use changes, you don't identify land use changes. Approach two, you identify land use changes. Approach three, you identify units subject to the land use change. A country may have one region that is approach one, another region which is approach two, another region which is approach three or various. Any kind of subdivision that you want to have, you may assign a different land representation to that. Okay, the area that I said you said. Now let's see how this is my country. I built the units of land. The software works in units of land. What is a unit of land? It's an area which is homogeneous by land use, current land use, land type, land user land type, and also historical land user land type. The data input shall begin from the first year of the inventory and go forward, not backward because the software automatically move the unit of land input like in 1990 to the following years of the time series, and accordingly move also from conversion to land use remaining. So far as the time period for the conversion is passed, then it is moved. You may select, let's go, sorry, I need to go a little bit faster. The area is where you select the approach for each, here you input each region. You may also do a single region for the entire country, of course, but you may go for different. You may have a project inside the inventory, as you wish. You input the area relevant for that subdivision and then you use the approach for the land representation that you wish to have for that one. This is a photo again, this screen here is quite small. Here I input the land representation for the year, which nobody see because we have this nice black, for the year 2025, so let's say the last years, you have all the units of land, you have a code to identify the land, you have the current and the previous land use type, you have the selection, I guess you need to trust me, but here this is the transition period, you may select the transition period for each unit of land. Here you input the year of conversion in case there is a conversion. Here you need to input the area and then you need to also show, tell to the software to which part of the time series this area apply. Here with this P you select, now I will go, just to be a little bit faster, I show you, there you select the methodology that you want to apply, pool by pool to estimate carbon store changes. So for each of them there is the IPCC default or the stock difference, you select which of the two you wish to apply. Here this is for instance the pool selections, you see, by default there is gain and loss because it's the IPCC default methodology, but you may move to the stock difference. This means that this unit of land, if I select the stock difference, this unit of land for the biomass pool will be shown in the worksheet for the biomass pool calculation, the worksheet that deals with the stock difference instead of the worksheet that deal with the gain and loss methodology. Here when I want, let's assume that input at the 900 hectare of area, then I click there, I select to which fraction of the inventory time series I want to apply, there are, by default is the current inventory year and forward because the data input need to be done from the first forward, but then you may decide to apply only for this year and you may decide to apply backward in case of corrosion, this functionality sometimes are needed, but the most commonly used is from the current inventory year forward. Okay, the software told you keep the land unit under conversion according with the selection of the area, or sorry, of the transition period, and then after the conversion period is passed, you see the software show you the unit of land in a blue color and then you may decide either to keep tracking it or to merge it with another unit of land of the same characteristic. These two make, I mean, easier than the data handling in the carbon stock estimates. This is an example where I have then I merged. Finally, the software also show you the land transition matrix for each year, for each you select which is the region that you want. Of course, this works for approach two and approach three because approach one has no land use conversion so there is no land transition matrix. This is an example of my data that I built when I built the exercise. If somebody you want, I may give you the database, the ground data you may then check by yourself. Now we estimate, okay, I need to skip a bit. You see, the follow sector now that we are going to add some additional wash sheets is composed by 500 wash sheets, 502. This again doesn't mean that you need to fill a 502 but means that there is enough flexibility to allow to implement for each carbon pool almost each, methodological tier. I skip because it takes time. This is an example of the wash sheet to calculate the volume of stops changes. This is again, this is about harvest. This is about dome. Sorry. For the dome, like for the sun pool, you have again, you have a transition period. So you move it from one stock level to another stock level and then you input the transition period. This is the result. Now the comparison, this is a complicated unit by unit of land, by unit of land compilation that I did, but what is more important, we go to the last type of this one. This is not in the piece inventory software. I told you that piece inventory software ends with this calculation of emissions and removal stop, but I use, I compare it to databases. I have the result of two database for each pool. R is the reference. A is the actual emissions and the removal according with the actual data. So for the two years that are 2020, 2025 where I gave both the projection and the actual estimates, I made the calculation. Here I made a comparison and I hope I'm not sure you see, but if you see the last row, in the last row you see that always I've been very good in implementing my mitigation policies. The total carbon stock change is smaller. There is a lower loss. Still my country is losing. It's a problem of forest, the forestage, forest degradation. So the forest sector in my country is losing carbon. You see, old number are red and negative. I'm losing carbon. However, in my actual estimates, the loss of carbon is much significantly slower because my policies are going well. So compiling two database, one for the reference, and compiling outside of the software. The results give you an estimate of the performance of your mitigation actions according with the Greenhouse Gas Inventory classification, IPCC methodologies and so on. We're to download the software. The software is free of use. It can be downloaded in our website. Again, this is version 285. Another version will be released next month. We are going to release version quite frequently. But all versions generally maintain consistency in the dataset. So you don't need to input the dataset each and every time because it would be not welcome. The software has two installation package, one for the 32-bit, the other for the 64. You need to carefully select which of the two you need to install on your computer because otherwise it doesn't work. We are very happy to receive feedback by user on the software. There is an email, you write and we respond very, very efficiently. Thank you. This is the office where we are. Thanks, Sandra. It's working. Yeah. Yeah. Yeah. Yeah. Okay. That's better. I see. Yeah. Take a breath. Okay. Thanks, Sandra. Thank you. Yeah. Thank you. Thank you. It's a great combination with Inga around the wall to wall add-on we have been working on. This, as Sandra mentioned at the start, there's two add-ons being produced. I'm going to present the one that's being produced about what the changes are going to be and how they're going to be implemented in the software. What are the characteristics of the data that we're trying to generate on C-Pool? So really the add-ons that we're producing are designed to support data collection for consistent land representation. This was identified as a challenge for inventory compilers to utilise the IPCC so they're really designed these add-ons to enable a bit more optional functionality where possible as well so where countries can tailor the data to their own specific needs including subcategories of land and then also obviously to facilitate a more seamless streamlined capacity to upload into the IPCC software so that we don't have to manually input the information seen from Sandra's presentation some of the data sets can get quite large. So another key point on that slide is that these add-ons will be obviously voluntary to be used it's not a requirement to use the software with the add-on, the add-on is an optional add-on basically. So the main trigger here for this particular add-ons are to ensure consistent representation of lands across all land uses so we're not just talking about forest non-forest, we're just talking about the six land use categories within the IPCC guidelines so in this particular add-on the wall-to-wall add-on in SIPL we're looking at image analysis for land classification at the pixel by pixel we're tracking that land through the time series in a consistent way correcting for bias in line with the IPCC requirements filling within the time series as well and we'll talk about how the algorithm we're using is doing that and compiling a land representation data set as I said it's an export file that can be uploaded straight into the IPCC software and that will be unit of land by unit of land so Is it fine on zoom? It's good, sorry about that if you're on on zoom So the process we're taking in the SIPL platform is to infer data on land cover and land use using back here, sorry, I'm deafening you as well technically incompetent so yeah using multi-spectral and multi-temporal analysis then we're up the technique is really requires the upload of a baseline map to initiate the classification with the algorithm we're using so that can be a national specific map or a global map so that we just basically you need an initial baseline map to start the process and then the algorithm in this case CCDC can classify the pixels into the six IPCC land categories and most importantly the transition between those through time so at the moment we're demonstrating with CCDC but there are other the capacity to put up any other algorithm is also fine as long as the key criteria is that you can classify the six IPCC land classes through time which you know you should be able to do with a good base map now I can't change the slide there we go so the process that we're going through will provide some logical rules to aggregate and classify the pixels and that will be in accordance with IPCC land category definitions it allows as I said the addition of user specific subcategories because we all know there's many many different subcategories under the forest land category and also possibly cropland and others and then helps to enable that delineation of unmanaged land which we saw from Sandra's presentation in those forms is a key requirement as well for classifying land so then we will stratify the pixels through the time series by these subcategories and enable users to put in their own transition periods but the default will be that 20 year transition period so there's still the functionality of the software will retain that you can make country specific changes on the transition period as well and then they'll be exported as these annual matrices either as approach 2 or approach 3 data set so that's the I guess the main features and functionality this part obviously is quite challenging for us at the moment in the approach 3 frame but we're still working on some process for estimating and eliminating bias and documenting that and we're doing that in collaboration with the G4I MGD component as well so we will produce a method for doing this and then also enable the user or the compiler to extract the data and apply their own techniques and this is important to keep that functionality and the country specific nature of the process and then gap filling of course it's going to be built in accordance with the 2019 refinement and more advice was provided in the refinement and again this will have optional functionality where the country can actually remove the data at that point to apply their own method for gap filling as well and then the final step of course is to compile all this into the time series for each land unit and then export in a format that can be easily uploaded straight into the software so that's really the technical specification of the add-on that we're building in SEPL and I'm going to hand over to now to present some nice graphics before I do that sorry there's one more slide about when will it be ready because I'm sure you're going to ask that question at the end so I'll pre-empt that one by saying we're aiming to release it in the third quarter of this year so quite a bit of work has been done we're working through the documentation and the final tweaking of it to make sure in testing and yeah definitely if you're interested in testing it as part of the community we'd be really welcome some reviewers or testers as well so you can just let us know at the end and we'll put you on the list and we'll see you again. Thanks Cardi. Yes thanks a lot so actually you have the challenging task just to show you how it will work because I said it's not ready yet so thanks Karls for indicating this so SEPL I think all this community knows of it you heard a lot in the last date so in case you don't know please come and see me or Eric so all these slides are courtesy of Eric Lindquist who's the SEPL technical officer so SEPL stands for the system of land monitoring so this is also why this add-on has a place here we do have also the developing team in the room so you have also technical questions on this so the idea is so the SEPL platform for the ones who have not having access so it's an open source cloud processing platform so please feel free also to go and look it out so as Cardi was saying it's CCDC based so it's a continuous change detection and classification the nice thing is SEPL was also made for this it could be used by dummy so people are not familiar with this geospatial analysis although through all the sheets we've seen from Sandro it's a bit more complicated than just for dummy so the good thing is that you can use any area of interest any time periods we have heard on the default of course there's about a 20 year period but in theory through this module you could use any time periods any sensor any sensor combination for the ones who using the SEPL platform already so you know there's a mirror of the lens at the Sentinel and there's a planet data so they can be used any dates combination as well which is also nice for the transitions of the land representation so as Carly also was saying so you need of course a map to start could be a national map or global so this is an example of Uganda and you need of course the classification of a start time for the classification so then if you choose a start time you can apply the different classification models so we heard the CCD but it's of course not limited only to this at any date which can already verify with separate change detection so you can remap and unlikely transitions you can also see which does pop up so more to come so as said hopefully ready soon by Julian thank you now the sample buys the tool is not showing there with ok I don't know how to do this can we just wait wait wait this one this one this one ok you see it there it's too complicated now wait wait back and here sorry too many ok now I think we see everything the same finally sorry now so you choose this one ok sorry here we want to present you what we have realized in collaboration with the PCC task force in ASCAS inventory for Collector to make this software now integrated in the software as you know Collector is the software that has been already used a lot from many countries to do their land representation using a sampling approach but I want to emphasize that starting from 2024 under the transparency framework of the climate change convention and I see many people here from developing country there will be a real verification also for developing country and things will change and the real verification will be on the clean ASCAS inventory so we highly recommend that most countries start to think about using the PCC software to generate the inventory and if they want to use Collector we will facilitate the integration of the data that they are collecting we collected in the PCC software so this is just to introduce you Collector that as you know it's based mainly on Google technology but not only we also take other open source data it's something that we have developed looking to what the large large majority of the annex one country before under the Kyoto protocol and the convention were doing to report the data and the convention so this tool was developed exactly to enable any country is the case of Niger where we have been just two weeks ago to replicate what the country like Italy like Japan US and so on many other are normally doing so that use point instead to have a cartographic approach to generate the statistics and when you do this point approach let's say you may focus much more on a single point and you may go really much more in detail so for example here we don't classify the six IPCC category we go at the detail subdivision of each country so sometimes we have classification depending from which country we are talking about in Mongolia we have 12 different type of grassland because in Mongolia grassland is super important in other country where other category are relevant the land use category you may have different type of things so the idea is that okay you build these land use time series you always start from the present in collected and you go back using the geospatial data the satellite data and you try to track if there has been changes for example here we have done an example where from settlement we pass through cropland and then to forest and these data the operator have to our spirit is to to see these from position from the point of view of the country of the people of the country who has to do this work so we try to simplify their job so if there is a change then they will have to pay attention but if there is no changes the data completion is much more simplified so we have and now we have an experience of 10 years of developing of collected and more than 40-50 country that have used it and we have not only used collected for these type of things now collected has been used also for hundreds of thousands of research activity we have many many scientific work done but let's say our point of view is try to simplify the work that the people has to do to do the land representation to report under the convention and we have developed now in the software when you download the software each user has already the option to adopt what we call a CHEP file with a collector project which already has as a background all the data arrangement ready for the integration of the PCC software the integration with the PCC software is not a joke I don't know if you understood from Sandro presentation that there are hundreds if not thousands of combination but the user don't care about it this come automatically by this template that we have prepared on top of this template the user can build whatever they want but as a default all the potential combination between vegetation climate soil time and so on are there and this has been already done developed by Alfonso and you just download it's your choice if you're doing for some type of research that has nothing to do with carbon PCC you don't care but if you are collecting data you would like to integrate one day this with the PCC software we suggest that you start your development of your project from this new template that is now available from our website you will see after you have done your compilation now for example Niger is because it's just one of the last country where we have been has already used this template and once they will have collected their data there will be a menu there is already a menu option for them to transfer the data directly in a format that could be directly ingested by the PCC software so once the data is important and a PCC software all the land representation part which is many of these hundreds table that Sandra was presenting will be automatically done automatically done by just this import of this data set realized by with collector and you will have all your matrix okay here we should have put many years in the future and you will have for the moment now we have the possibility to have all the period where the assessment is being done but in the future we are going to put also the back casting and other element like for example the assessment of the uncertainty and so that for the moment we have not done but we will work immediately after but for what we have done is that now any person who wants to use collector to do the land representation can from today use this export function with the PCC software and Alfonso now will show you we will do a real demo so we will have an example and we will show you how it works I want to say something like which is really important when you use collector and use the sampling approach for example now we are doing Italy let's say you will have your basic unit that Sandro was showing can go up to your single point so in Italy we have 300,000 points 300,000 points and this theoretically can be aligned in your green ask us inventory then there are some aggregation which simplified it and 300,000 it's a lot generally we are around 30,000 50,000 point if you go world to world you may be in the dimension of millions because then as they said it's pixel based approach so it's easy that every possible combination then becomes aligned so we have always to think about the experience of the end user so the end user here is the person who has to do the national green ask us inventory who will be verified really verified like has been done now for the next one party under the convention and it's not going to be a funny work but someone has to do this work so we are trying to do whatever we can to simplify the life of these people because at the end the performance of each country will be assessed with this tool with the inventory the national invent so Alfonso please go ahead and now we will do just an example with a data set that Alfonso has prepared okay so just very fast first of all if you have data that you have already collected with our usual IPCC template because we have been using this template for years and whoever is not using this template shame on you because we have been telling you to use this template you've done really bad so we still can use that data you have collected from the previous template now we have an improved template that I would recommend everyone that is going to make a new assessment to use it it's already on collect on the survey designer so you have just collected you know that you designed the survey in collect and in collect you can already when you go to the new survey we have this collector plus IPCC survey so you start from there you are good don't mess with the IPCC attributes of the survey please alright so we have an example here of some data that was collected in Benin it's just a few thousand plots that were that I'm going to use as an example where we have the usual IPCC that was okay everybody knows Benin anyway so you have your usual LandCover TreeCounts RangeCategories and as you know you can classify each plot with the current land use the previous land use and if there has been a third land use before also that the next improvements will be that we can have unlimited land use changes but up to now we are limiting to three which basically covers 99.9% of the cases that we have so far anyway so when you have collected your thousands of plots which takes you some weeks you are happy with that you review you clean you do everything so you will aggregate all the collectors data as you know we have the interpreters the interpreters of the interpreting they send you the data you put it into a single database and you have here in tools a very long option here where you can generate the GIG package just to make this faster because this takes a while to generate the database I'm going to use the last one that I have generated the first thing is you have to set here your country this is not like an amazing interface yet but at least it's kind of working this is important you can select how do you disaggregate the data because in the IPCC software you can have data at country level but you could have it at province level or whatever level you want so in this case for instance for Benin we could have it at national park level for instance because the data was collected for national parks but let's say that we just have the province level okay next you have a lot of questions so you have to define for each subdivision if it's manager or manage and what type of for instance what type of trees in the forest if it's a crop land again if it's annual or perennial what type of perennial if it's land grassland same if it's manager or manager or a manager or what type of what type of grassland okay because this is requirements for later when you import the data in the in the other in the GICI tool but Benin you know manager or manager and at the end you click finish you say you want to generate a file it generates a zip file it's a zip file okay this is probably going to take a while so I just got to the one that I just generated before so this is the this is the file that it generates it generates so it's not going to fail I'm not lying it will generate the zip file it just takes a while alright so you get a few files and these few files are quite interesting here because it automatically will generate the land use matrix okay so for each year if I open this it's already open okay so this is going to generate the year by year land use matrix alright it doesn't doesn't show it I don't know what happens there I think my excel just got too much on the plate so because we are pretty large files or with a lot of stuff so eh I wanted to show this to you because this is interesting because basically what we have done in collect you can do this what we call automatic sorry we can do this calculate that attributes so we can go year by year thanks to the for each land use change we collect the year of the land use change so we can recreate for each plot the timeline year by year we know each land use change has happened so we can say okay even if we didn't in the form that the land use in 2010 was forest since we know that it was forest in 2022 and it didn't change the whole time so we know in 2010 it was forest it was cropland in 2020 and we knew that it had changed in 2015 from forest to cropland we know that in 2010 was forest also so we can reproduce the whole timeline so as I said here you can create the whole land use change matrix for every combination of years okay so this is 2000-2001 2002 everything we can also do the same by stratification which you get a huge table with all these land use change matrix by strata divided by climate, soil and ecological zone which is a bit useless because it's just so much data and then finally we have a lot more stuff here that I'm going to show you because it's really big look this for instance the whole land use change matrix for each plot which also can be useful and you are going to do analysis outside of the greenhouse gas inventory tool maybe you are going to use R or you have your own process or whatever so you can take this file and import it in your R process for instance you are into R, I see some R people here so you can get your whole timeline with the category and areas for each plot so that's the final thing you go here now when you are happy with everything you can go into the greenhouse gas inventory tool you have this export import functionality import worksheet data here you could choose then the the file that we created so this is the one that you import the one called GHGI tool data ok so you will import that one and again I'm not going to do now live because it takes a couple of minutes but I just imported this before again I'm not lying I just did it before so it's fine once you import this file you go to your land use manager and you automatically have filled all the different combinations of land uses ok in the IPCC software it's a bit complex because you need for each land use subdivision you have the combinations of the soils climates and ecosystems ok so you get all of these which will take you a while to do you are very happy about that you close that then you go to land use manager here you see that we have the data not by the country of Benin but by the regions so you have the regions of Benin so you could do your greenhouse gas inventory at province level national parks or red projects for instance you have like a red project division you could export the data at the red project and then do the inventory at the red project level ok again so I have my regions I'm sorry probably you don't see this very well then you have for each region your land representation so if I change here for this one for instance this would be my land representation there is a back here so actually it misses 4000 from the province but anyway who cares it's just a test you can go now to your forest you see the land units that Sandro was explaining before so you have for each land unit if it has changed or not so this first one is all of the forest that has remained forest and here for instance you see some random strange plots that have changed from cropland to forest ok and how much is that and finally you have the matrix and in the matrix this will be the matrix for the current inventory year so this is the land the land representation matrix for each region more or less what we have shown you before in the Excel format but now per region and just for you to review what is the changes that you have had in the province for each inventory year I have to say this is a bit annoying because you need to change the inventory year you need to go back here and change the inventory year somewhere like that so it's not super user friendly but the user friendliness of the software is an issue but we have to make so many upgrades that sooner or later something else will be changed thank you very much just just to add that both tools are still in a development phase so we didn't test anything we are happy for what they shown but there are still big pieces of functionalities to implement one clarification the unit of land is an area that is homogeneous so many plots may be aggregated in the same unit of land as many pixels can be aggregated so when you import you don't import each and every pixel into the render representation of each and every sample path depending on the variability of the landscape and the number maybe bigger or smaller not at the level of the million of course cannot be managed at that level of detail now questions Giulio no question from the floor I don't know how it works do we have a micro? thank you very much next thank you for this demonstration it's a lot of work that's going into this and it's quite impressive I have one sort of not technical question but conceptual question around the land use change matrix and using either sample based approaches or wall to wall land cover maps there are both sort of representing land cover change not land use change and manual step at the end of that process where users can go in and sort of correct the land use cover change matrix and for example in Italy there's forestry forest to grassland that is not a land land use change that's a harvest and so would you allow users to correct for those issues or not yet? the software is not an interpretation of the guidelines you need to apply the guidelines so in the for collectors the people that does the classification should be able to recognize if it's a deforestation or it's just an harvest because it's an expert basis in this other case also that depending on the how good is the algorithm should be able to do so so it is generally this is not a land cover change however why the land use drive which of the methodology I need to apply to estimate the carbon change and so on at the end is the land cover that tell me if the carbon is still there or not so if I see forest disappearing either I estimate this carbon loss as an harvest in forest land or as a deforestation in forest land converted to grassland but the amount of carbon that is going to be the same what follows is different in a case you should be the same sorry in the harvest scenario if you're not clear cutting it I'm assuming a clear cut I was assuming a clear cut replanting a very basic system so I know of course depending on which of the two types of action occur anything else that follows is different in one case the forest will grow in other case no so it's a big difference but for the instantaneous calculation of carbon is less Sandro there is a question from Brian Mutascia with the sampling based method how do we ensure that classes are representative following the systematic grid so as the slide shown by Danilo the density grid Danilo are you going to take that or I may say something again it depends on of course who will apply the sample based approach and need to have a sample size that is big enough to capture the entire variability of the landscape depending on how much is variable the landscape the sample size changes this is I mean let me clarify one point even the sampling design doesn't come with the collector the sampling design is something that has to be done by expert in the country with for sure one statistician who knows what is talking about and each country is different so collectors just ingest a sampling design that has been designed and built in order to do what Sandro was saying that should be able to consistent represent the country and be able to report with a good accuracy all uncertainty the key category the relative changes that are occurring in the country so each country has a different should have a different strategy if you go for just a systematic sampling approach without any stratification then most likely you are going to have high density of the sampling design. Anyhow the use of the tool does not replace the need of the expertise on those that use the tools otherwise you don't get anything sensible. Question? Thank you for the great work John the woman from Uganda I am whether it is we are uploading data from correct earth or maps I am wondering at what stage do we input biomass stocks into the software because ultimately you need the land use but you also need an estimate of the biomass stock the software has war sheets some biomass data are to be input when you characterize your land use type orders data are to be input into the war sheet calculations I have been very fast because of time and then any other screen is so small that you cannot see but if you look at the presentation I guess it will be available it is available. You download it you look at your screen and you will see how to populate for the carbon stock and then calculate the carbon stock changes. Next question just there is easier because you are close you will be the last. Thank you all presenters for your nice demonstration I think what we have been doing outside of this IPCC software like SEPAL and collectors now you integrated with this that is excellent so I have to appreciate for that. Having said that I have two questions one is about this land representation land maybe I don't know outside this out of these the IPCC has six classes but in our I am from Ethiopia when we are doing this change detection for some projects result based payment projects we received some comment actually that comment is very nice because one IPCC class which is grassland is categorized as they have pure grassland as well as the shrub land but these two land cover are completely different in terms of biomass or carbon maybe for the future I think for updating during updating this IPCC I think there are many people here which will be engaged in that update this is simply a comment to consider putting or taking these six into seven classes because they are completely different this is one of my comment the other is maybe collectors you integrated collectors but these days I think we are using we are extensively applying these collectors online maybe it is more easier as I saw it than collectors everything the design the survey questions many things are more advanced I think can we maybe do like the add-ons like collectors like collectors online instead of collectors thank you first the subdivisions are to be input by the user you may input a thousand of different grassland types it's up to you there are not pre-defined because IPCC has not pre-defined the land use subdivision for grassland but you may do any substratification that you wish about collectors then at the end but my first answer is that we are not going to to have an online tool with collectors so the tool will be the one developed by FAO it will be web based at the instruments thank you we are using the collector tool for many years now I think I don't know if I missed in the beginning but it would have been nice to see a timeline of where we are now and what the next steps are because BTR is submission by the end of next year we are still developing the tool and you would be nice to know your expectation of how long it will take for countries to uptake this tool it's very complex even we've had experiences where even the trainers had difficulty understanding the tool so it's something that would be nice to think about in terms of the BTR and then I was curious if the tool allows for if you have a sampling design like a systematic sampling design if it allows for the post stratification you know within the tool if you want to improve your estimates if you have a map and you can post stratify your systematic sample thank you then I give the floor to the relevant these activity has been strongly impacted by COVID and and the front so it's up to you I mean we collected the post stratification should be done by in the sampling design after you collect the data so it's nothing that you reconnected to the IPCC greenhouse inventory tool should be done outside of that through collectors again and then re-imported into the tool so right now there is no post stratification stepping in the tool in the greenhouse inventory tool sorry first what is your respect at the timeline because I say everything is under development the software will be available with this interoperability by the end of the year but not with the add-ons the add-ons is foul so I guess on my slide we're aiming to try and have it level by third or fourth quarter this year we have been working on it quite intensively so we just showed a small part of it today so that's our intent and I think for the uptake by the countries it will depend on the capacity building options we have specifically with the countries where we're working already so I think they will have a slight advantage on the other ones and for the new ones I think we will have to follow up a plan in this whole family of resources and see in the gap needs assessment together with the CLP which you have heard so I think that's going to be the two tracks regarding the collector as we say the tool will be available now the add-on and we are already testing some country country like Mozambique where we know that you have done several of these things for you there will be a transition you will have to do on integration and so on I should be something that can be done by this year next year so ready for the reporting of next year for sure but we really the real advantage will be the use of the IPCC software so we really encourage everyone to do this transition because from our point of view it's going to be even from the IPCC side the first software that will solve a lot of reporting problems then as soon you move it's better thank you very much I agree with Danilo that visual interpretation of samples is very meaningful because we can much more look into details on the other hand we know that most of these area changes are also rare events if we consider sampling so looking at points even if we use a very dense grid will result in very small numbers of changes that are affected by huge errors it can be it can mean all or nothing how do you account for these uncertainties in the area estimates let me say indeed there is a big fraction a big portion of the work to be done by both is the uncertainty the quality calculation and the bias correction which is not yet done the answer is that the moment when the work will be concluded possibly this is not a question to collector it is again a question of sampling design in the sense collector ingest a sampling design by here we have Marina from Italy they have the experience let's say Italy is not a country where you have big land use changes all the changes are rare event they have built an assessment with 300,000 point they did also some estimate some scientific paper where they say 100,000 could be enough based on what they have done but just to give you an example it's nothing to be collected but in a country like Italy where any changes it's a rare event because the territory is stratified since century this amount of point is enough to capture the details but nothing to do with collector please and tomorrow there is an event with a side event when they are going to approach this issue of creating grids that are more suitable for rare events let's call it the ensemble sampling design there is a side event tomorrow about that so yeah uncertainty I was trying to look at the tables the screen is too small etc if the IPCC will also deliver the estimates of uncertainty we apply also the uncertainty analysis at the moment it's limited to be further developed but you need to input the uncertainty value then it does just error propagation according to the IPCC that was my next question Carlo will be implemented in this just after interoperability the main task will be the uncertainty analysis too to be completed and also to include approach to Monte Carlo analysis another question from another remote question sorry for the pronunciation to help speed up developing countries ability to transition to the ETF are there plans to create a community of practice for developing countries who want to use the IPCC software plus supporting tools for the national reports it will be wonderful the question is my office is 3 persons so this is not our main task so yes of course we need to convince some institution FAO, GFOI somebody else whom we cooperate but we cannot physically do we are already doing by far too much there are physically so yes we are happy I mean we need the help of countries if countries decide to do this we are extremely happy to do because we are going to produce this guide book it is extremely complicated also to produce that because it is a big task just to replace the fact that we cannot give answer or train anybody but we hope that somebody else will do it right so we are not transitioning to anything new we are starting something very new and that is a transition for a new era for everyone also for developed countries what I think it will be very very interesting is that people get used to this and then exchange experience right, rather than trying to do capacity building on something that even the capacity building didn't tested yet I can offer maybe to manage this community to me just to manage the community a practitioner exceed my time availability because already answer emails sorry here there is a lot of Luka, Luka Fest and then there is Randy Jim Jim not everybody knows me so I think he needed to be guided to me but it is very helpful it is great so just a question regarding the application for Red Plus because some country in your example you saw that each Red Plus activity was mapped to land use change categories so some countries might not to do so or for example some activity can happen in the same land for example so how would you advise a country to address this issue for example I mean we do Greenhouse Gas Inventory my message is you do your Greenhouse Gas Inventory and then you include all activities by default so just the inventory you have forest land and many forest land from and to and you have everything that can be under red consistently estimated also you may even submit as they are to the UNFCC confusing for people here is these red labels right because they are not consistent and for example you can say something is forest management sustainable management of forest but at the same time it has a decrease in forest management stocks and someone else can classify that as a forest degradation so I think the point is to be transparent and explain very well what you say is happening and don't try please to start a process of defining things no go like Sandra said that's much easier this is proving the past to be a disaster we have to be very careful we have to see what are the changes report the changes label the changes as you wish and be very transparent and very open and that's it that's the best way to start maybe in 20 years you will be able to start thinking about other things hi I'm Chip Scott I'm with Silva Carbon so I guess a two part question first of all I understand that currently you have this summary tool to develop your estimates but there's no uncertainties no variance is calculated is that correct at the moment the tool is not completed so there are not uncertainties once completed there should be uncertainties that will happen by the end of by the end of the year is FAO I don't know so I guess my offer is that Ingo we've been talking about sharing the tool that Andy Lister and I built couple years ago for analyzing collect earth data so I could offer that to anybody who's interested it's an Excel spreadsheet happy to share it the other thing is Lutz and I were talking just earlier today about trying to do it he's already built a tool in Power BI if people are interested in working with that so just an offer the more we test our tool in different situations the better it will get thank you the last question to the gentleman here I don't know thank you very much my name is Kari from Indonesia in my understanding within the last three days and also from the software that has been described quickly very quickly for me that we talk much more about the above the ground biomass if I'm not mistaken we haven't go to the details about the carbon below the ground the first one I just have a quick reading on the IPCC guideline 2019 there has not been a specific equation or formula which is directly correlated between the ground water level and also the CO2 emission because it does come to the challenge for us in Indonesia they have a huge work that has been done but has not been incorporated in the calculation so that I wonder or maybe that I'm encouraging all of us to see that in much more details because I'm working in the pitland area it's different issues on the carbon storage because you know exactly that the three is here it's very different with the pitland area and we have a huge pitland area so as I say that maybe we put the formula in the IPCC panels the first one and maybe we will develop because we have the soil moisture data even though that we don't know exactly what's the correlation with the soil moisture data and also the ground water level that's my comment thank you there is the wetland supplements that has some default for sure can be improved also because it's from 2013 and of course country may require IPCC to do further work to refine also the wetland supplement the software you have already the wetland supplement so the equations are there those available for IPCC also to estimate the pitland losses it's not a criticism on collect earth in your example on Benin you were very lucky to have data until 2023 unfortunately in the project areas where I try to apply it I sometimes find 5 year old data in one part of the area and 3 year old data in another part and completely unsuitable data in another area what are your experience in different countries about the timely availability of imagery and reporting based on visual interpretation ok when we train people using collect earth we always explain that the very high resolution images should help the local expert to understand the context the geospatial dimension if there are many very high resolution images it's definitely much better and the operator can just use this to do everything but always you always have starting from 2000 the possibility to check Lancet and most recently the Sentinel up to 210 meter data so the expert and it's some years ago was was not so in some part of the world I can tell you in Papua New Guinea we were we had a lot of trouble because in Google Earth there were no problem where the people from Papua New Guinea who told us please use Bing maps we are using Bing maps to see very high resolution images because Bing maps did already the full cover of the country while Google didn't so the fact that Bing map is one of the input for data in a collected is coming from a suggestion from Papua New Guinea people so now I can tell you there is probably no part of the world where there is no one very high resolution images often there are many but at least one so that the people understand the context and then use the let's say the medium high resolution image of Lancet or Sentinel to see the time series we have solved in this way in the tropical countries we also have the integration with planet maps this planet imagery that should be even one month old okay this time we are very short we are out of time it's very short thank you I'm only from Indonesia very short and quick question thank you Alfonso for quick demonstration my question is how about detail scale I mean detail scale that you have projects for rain cover map is it come from the Lancet imagery or some kind of imagery that you use or we can enhance if we want to enhance our detail information maybe using the confusion imagery or something like that thank you for Alfonso I think both sides they said yeah how about the scale level that you have produced rain cover map we don't produce rain cover maps in earth map so collector is for just data collection sampling based data collection and we even discourage the use of rain cover maps okay I guess we may close I will close as IPCC then I leave the floor to GFI so as IPCC TFI we provide a specification for developing these tools our cycle IPCC works in cycle so IPCC will keep developing the software in the next months but this activity with FFO at the moment is to be considered I mean because of the action changes are going to happen we don't be working on this as IPCC so maybe GFI may take the lead to keep working on this or following the work on this tool according with the specification I hope that IPCC in the new cycle will have further consideration of the tool and possibly a further action in this but the software is going to be developed with no interaction don't worry thank you very much GFI I just want to back up Sandro's point of view definitely the MGD component has these add ons and the continuation of those on their work plan and I really enjoyed the presentation about how we as a unit as a collective GFI to support countries to utilise the IPCC software so I think that's something we can take back to the office and work out with the partners how we do that about the community of practitioners this is an idea I thank the Marcela but this is an idea that is already considered and the message will be conveyed by us to the next Bureau of IPCC to consider this opportunity